knitr::opts_chunk$set( collapse = TRUE, comment = "#>", fig.path = "man/figures/README-", out.width = "100%" )
1. Install Giotto python Environment within R
In terms of Giotto Suite, the best way to install and load Giotto python Environment is to use the installGiottoEnvironment() function and activate the Giotto Environment via createGiottoInstructions(). The installGiottoEnvironment function will install required packages from r-miniconda and createGiottoInstructions will search for giotto environment and load it into R.
library(Giotto) installGiottoEnvironment() instrs = createGiottoInstructions(save_dir = results_folder, save_plot = TRUE, show_plot = FALSE)
2. Manually installation of python Environment with conda.
Unfortunately, sometimes r-miniconda may have a problem or something else happens. A manually installation of required python package via miniconda is needed. Before you try to install python packages, make sure you have the miniconda installed
2.1 Create conda environment and install required packages
python
conda env create -n giotto
conda activate giotto
conda install pandas networkx python-igraph leidenalg python-louvain python.app scikit-learn -c conda-forge
2.2 Check the conda environment location
conda info
You should be able to find the conda environment path/to/conda/envs/giotto from the listed active env location
2.3 Determine the path to this environment when you create the Giotto instructions or Giotto object. Please double check the name of the path. Some MacOS system has the same folder structure as linux.
for OSX:
path/to/conda/envs/giotto/bin/pythonw
for linux:
path/to/conda/envs/giotto/bin/python
for windows:
path/to/conda/envs/giotto/python.exe
2.4 load the python path to R session. Note you need to change the path to the actual location of the installed python environment
library(Giotto) intrs = createGiottoInstructions(python_path = "path/to/conda/envs/giotto/bin/python")
Alternatively, load the python path directly from reticulate would also work:
reticulate::use_python("path/to/conda/envs/giotto/bin/python")
Or this should also work: note the conda_install_location path should be able to found by "conda av data dir" from conda info
reticulate::use_condaenv(conda = conda_install_location, condaenv = "giotto")
Note that load library(reticulate) might result in loading the first python environment on the machine and the loaded python cannot be overwritten until a new R session was created. So avoid loading the reticulate library before a giotto python environment is loaded. You can always check if the correct python was loaded:
reticulate::py_config()
